检索结果-内蒙古大学图书馆

A hypothetical comparative evaluationsystem for arctic indoors

Frontiers of Architectural Research 2025年第1期14卷 210-223页

作者： Seyed-Amin Tabatabaeifard Jean-Francois Lalonde Marc Hebert Andre Potvin Claude MH.Demers GRAP(Groupe de Recherche en Ambiances Physiques) Ecole d’architectureUniversite LavalQuebecCanada Computer Vision and Systems Lab Department of Electrical and Computer EngineeringUniversite LavalQuebecCanada CERVO Brain Research Centre Faculty of MedicineUniversite LavalQuebecCanada

This research presents an innovative approach to evaluating indoor spaces,combining qualitative attributes with numerical architectural metrics. A hypothetical comparative visualization system is introduced, utilizing HDR visual imaging and thermal imaging in 360° field of view across multiple indoor environments. The study aims to provide architectsand occupants with a user-friendly tool informing them about the primary considerations oftheir built spaces, with a specific focus on indoor environmental qualities in remote Arctic regions. Key inquiries delve into the efficacy of the spherical approach and the capacity ofcomparative visualization to offer insights into space quality. Preliminary experiments contrastindoor environments in terms of circadian lighting, thermal uniformity, and view access tooutside in the 360° field of view (VAR360). The resulting visualizations hold significance inintroducing an immersive approach for depicting specific non-visible environmental qualities,particularly in relation to the window characteristics of spaces. It demonstrates the integration of multiple environmental variables, both steady-state and temporal, from central pointswithin spaces, providing a comprehensive view over their non-visible qualities. These resultsshould be useful for researchers and practitioners within building sciences, computer vision,and photobiology, showcasing an out-of-the-box approach for categorizing indoor spaces basedon standards and human-environmental qualifications.

关键词： Environmental assessment 360°imaging Comparative systems Low-cost tools Built environments

来源：评论

学校读者我要写书评

暂无评论

HyperSteiner: Computing Heuristic Hyperbolic Steiner Minimal Trees

HyperSteiner: Computing Heuristic Hyperbolic Steiner Minimal...

引用

2025 SIAM Symposium on Algorithm engineering and Experiments, ALENEX 2025

作者： García-Castellanos, Alejandro Medbouhi, Aniss Aiman Marchetti, Giovanni Luca Bekkers, Erik J. Kragic, Danica Amsterdam Machine Learning Lab University of Amsterdam Netherlands School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Sweden School of Engineering Sciences KTH Royal Institute of Technology Sweden

ISBN: (纸本)9798331311995

We propose HyperSteiner – an efficient heuristic algorithm for computing Steiner minimal trees in the hyperbolic space. HyperSteiner extends the Euclidean Smith-Lee-Liebman algorithm, which is grounded in a divide-and-conquer approach involving the Delaunay triangulation. The central idea is rephrasing Steiner tree problems with three terminals as a system of equations in the Klein-Beltrami model. Motivated by the fact that hyperbolic geometry is well-suited for representing hierarchies, we explore applications to hierarchy discovery in data. Results show that HyperSteiner infers more realistic hierarchies than the Minimum Spanning Tree and is more scalable to large datasets than Neighbor Joining. Copyright © 2025 by SIAM.

关键词： Trees (mathematics)

来源：评论

学校读者我要写书评

暂无评论

Converging to a Lingua Franca: Evolution of Linguistic Regions and Semantics Alignment in Multilingual Large Language Models 31

Converging to a Lingua Franca: Evolution of Linguistic Regio...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Zeng, Hongchuan Han, Senyu Chen, Lu Yu, Kai X-LANCE Lab Department of Computer Science and Engineering MoE Key Lab of Artificial Intelligence SJTU AI Institute Shanghai Jiao Tong University Shanghai China Suzhou Laboratory Suzhou China

ISBN: (纸本)9798891761964

Large language models (LLMs) have demonstrated remarkable performance, particularly in multilingual contexts. While recent studies suggest that LLMs can transfer skills learned in one language to others, the internal mechanisms behind this ability remain unclear. We observed that the neuron activation patterns of LLMs exhibit similarities when processing the same language, revealing the existence and location of key linguistic regions. Additionally, we found that neuron activation patterns are similar when processing sentences with the same semantic meaning in different languages. This indicates that LLMs map semantically identical inputs from different languages into a "Lingua Franca", a common semantic latent space that allows for consistent processing across languages. This semantic alignment becomes more pronounced with training and increased model size, resulting in a more language-agnostic activation pattern. Moreover, we found that key linguistic neurons are concentrated in the first and last layers of LLMs, becoming denser in the first layers as training progresses. Experiments on BLOOM and LLaMA2 support these findings, highlighting the structural evolution of multilingual LLMs during training and scaling up. This paper provides insights into the internal workings of LLMs, offering a foundation for future improvements in their cross-lingual capabilities. The codes are available at: https://***/X-LANCE/LinguaFranca. © 2025 Association for Computational Linguistics.

关键词： Neurons

来源：评论

学校读者我要写书评

暂无评论

From Generalist to Specialist: A Survey of Large Language Models for Chemistry 31

From Generalist to Specialist: A Survey of Large Language Mo...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Han, Yang Wan, Ziping Chen, Lu Yu, Kai Chen, Xin X-LANCE Lab Department of Computer Science and Engineering MoE Key Lab of Artificial Intelligence SJTU AI Institute Shanghai Jiao Tong University Shanghai China Suzhou Laboratory Suzhou China

ISBN: (纸本)9798891761964

Large Language Models (LLMs) have significantly transformed our daily life and established a new paradigm in natural language processing (NLP). However, the predominant pretraining of LLMs on extensive web-based texts remains insufficient for advanced scientific discovery, particularly in chemistry. The scarcity of specialized chemistry data, coupled with the complexity of multi-modal data such as 2D graph, 3D structure and spectrum, present distinct challenges. Although several studies have reviewed Pretrained Language Models (PLMs) in chemistry, there is a conspicuous absence of a systematic survey specifically focused on chemistry-oriented LLMs. In this paper, we outline methodologies for incorporating domain-specific chemistry knowledge and multi-modal information into LLMs, we also conceptualize chemistry LLMs as agents using chemistry tools and investigate their potential to accelerate scientific research. Additionally, we conclude the existing benchmarks to evaluate chemistry ability of LLMs. Finally, we critically examine the current challenges and identify promising directions for future research. Through this comprehensive survey, we aim to assist researchers in staying at the forefront of developments in chemistry LLMs and to inspire innovative applications in the field. © 2025 Association for Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Collaborative Deadline-sensitive Multi-task Offloading in Vehicular-Cloud Networks 17

Collaborative Deadline-sensitive Multi-task Offloading in Ve...

引用

17th International Conference on COMmunication Systems and NETworkS, COMSNETS 2025

作者： Kumar, Prasanna Sushma, S.A. Chandrasekaran, K. Addya, Sourav Kanti National Institute of Technology Karnataka Cloud and Smart System Services Lab Department of Computer Science and Engineering Surathkal India Siddaganga Institute of Technology Cloud and Smart System Services Lab Tumakuru India

ISBN: (纸本)9798331531195

With the growing technological advancements in the Internet and advanced functionalities in vehicular networks, it becomes crucial to execute tasks quickly and efficiently. However, the limited onboard computational capacity and vehicle mobility make it challenging to accomplish latency-sensitive tasks efficiently. Task offloading provides a promising solution to overcome these challenges. Cloud data centers provide efficient solutions, but returning the results to the vehicles takes longer due to the large physical distance. Leveraging edge servers to execute latency-sensitive tasks provides a fast, interactive response and less transmission cost. However, in a dynamic network, vehicles will be in constant motion with varying speeds, resulting in frequent handoffs from one base station to another. Our proposed work aims to select the optimal nodes to perform binary offloading with minimum cost using the collaborative vehicular network. We use a greedy-based offloading approach to address these challenges and achieve better quality-of-service and quality-of-experience in a dynamic environment to minimize costs, delay reduction ratio, and satisfaction ratio. The proposed work outperforms the baseline by 60.44%, 53.43% in reducing total system cost, delay reduction ratio, and 36% improvement in the satisfaction ratio compared to baseline algorithms. © 2025 IEEE.

关键词： Quality of service

来源：评论

学校读者我要写书评

暂无评论

Unveiling cognitive processes in digital reading through behavioural cues: A hybrid intelligence (HI) approach

引用

British Journal of Educational Technology 2025年第2期56卷 678-711页

作者： Lee, Yoon Migut, Gosia Specht, Marcus Industrial Design Graduate School of Design University of Seoul Seoul Korea Republic of Research Lab Faculty of Education and Psychology University of Oulu Oulu Finland Faculty of Electrical Engineering Mathematics and Computer Science Delft University of Technology Delft Netherlands

Learner behaviours often provide critical clues about learners' cognitive processes. However, the capacity of human intelligence to comprehend and intervene in learners' cognitive processes is often constrained by the subjective nature of human evaluation and the challenges of maintaining consistency and scalability. The recent widespread AI technology has been applied to learning analytics (LA), aiming at a more accurate, consistent and scalable understanding of learning to compensate for challenges that human intelligence faces. However, machine intelligence has been criticized for lacking contextual understanding and difficulties dealing with complex human emotions and social cues. In this work, we aim to understand learners' internal cognitive processes based on the external behavioural cues of learners in a digital reading context, using a hybrid intelligence (HI) approach, bridging human and machine intelligence. Based on the behavioural frameworks and the insights from human experts, we scope specific behavioural cues that are known to be relevant to learners' attention regulation, which is highly relevant for learners' cognitive processes. We utilize the public WEDAR dataset with 30 subjects' video data, behaviour annotation and pre–post tests on multiple choice and summarization tasks. We apply the explainable AI (XAI) approach to train the machine learning model so that human evaluators can also understand which behavioural features were essential for predicting the usage of the cognitive processes (ie, higher-order thinking skills [HOTS] and lower-order thinking skills [LOTS]) of learners, providing insights for the next-round feature engineering and intervention design. The result indicates that the dominant use of attention regulation behaviours is a reliable indicator of low use of LOTS with 79.33% prediction accuracy, while reading speed is a valuable indicator for predicting the overall usage of HOTS and LOTS, ranging from 60.66% to 78.66% accuracy,

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

XPM: eXplainable Process Mining 3rd

XPM: eXplainable Process Mining

引用

3rd International Conference on Advances in Information and Communication Technology, ICTA 2024

作者： Kim, Kwanghoon Pio CCSRI Kyonggi University Suwon Korea Republic of Data. Process Engineering Research Lab. Div. of AI Computer Science and Engineering Kyonggi University Gyeonggi Suwon16227 Korea Republic of

ISBN: (纸本)9783031809422

This keynote speech is to all intents and purposes introducing a new process mining approach and its implemented system, which are named as eXplainable process mining (XPM) approach and system, respectively. Through the XPM approach, we can selectively amalgamate a group of process event log traces, and through the XPM system, we can ultimately discover a series of complete and safe process models from a process enactment event log dataset. The XPM approach is theoretically based on the typical process modeling methodology of structured information control nets (abbreviated as SICN) and its algorithmic functionality is substantiated by the explainable ρ-Algorithm that is extended from the original ρ-Algorithm [6]. This keynote speech is also explicating the proven process mining algorithm that is able to discover the underlying SICN-oriented process model from a dataset of process enactment event logs and histories. More specifically, the concrete goal of this keynote speech is about an explainable way of discovering the complete structural formation of SICN-oriented process model. Assume that the structural formation is made up of an arbitrary number of combinational building blocks of the primitive process patterns such as linear (sequential), disjunctive (exclusive-OR), conjunctive (parallel-AND), and repetitive (iterative-LOOP) process patterns. Conclusively, it would be emphasized that the XPM approach and system ought to be very effective and well-fitted for managing life-cycles of very large-scale process models deployed in process-aware enterprises and organizations. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Information management

来源：评论

学校读者我要写书评

暂无评论

From Superficial to Deep: Integrating External Knowledge for Follow-up Question Generation Using Knowledge Graph and LLM 31

From Superficial to Deep: Integrating External Knowledge for...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Liu, Jianyu Huang, Yi Bi, Sheng Feng, Junlan Qi, Guilin School of Computer Science and Engineering Southeast University China Law and Innovation Lab Law School Southeast University China China Mobile Research Institute China

ISBN: (纸本)9798891761964

In a conversational system, dynamically generating follow-up questions based on context can help users explore information and provide a better user experience. Humans are usually able to ask questions that involve some general life knowledge and demonstrate higher order cognitive skills. However, the questions generated by existing methods are often limited to shallow contextual questions that are uninspiring and have a large gap to the human level. In this paper, we propose a three-stage external knowledge-enhanced follow-up question generation method, which generates questions by identifying contextual topics, constructing a knowledge graph (KG) online, and finally combining these with a large language model to generate the final question. The model generates information-rich and exploratory followup questions by introducing external common sense knowledge and performing a knowledge fusion operation. Experiments show that compared to baseline models, our method generates questions that are more informative and closer to human questioning levels while maintaining contextual relevance. © 2025 Association for Computational Linguistics.

关键词： Knowledge graph

来源：评论

学校读者我要写书评

暂无评论

Robotic Path Re-Planning for US Reconstruction of the Spine

IEEE Transactions on Medical Robotics and Bionics

引用

IEEE Transactions on Medical Robotics and Bionics 2025年第2期7卷 755-767页

作者： Assche, Kaat Van Li, Ruixuan Davoodi, Ayoob Tummers, Matthias Ourak, Mouloud Borghesan, Gianni Cavalcanti, Nicola Funstahl, Philipp Poorten, Emmanuel Vander KU Leuven Robot-Assisted Surgery group Department of Mechanical Engineering Leuven Belgium Flanders Make Core Lab ROB Leuven Belgium Balgrist University Hospital University of Zurich Research in Orthopedic Computer Science Group Zurich Switzerland

Robotic ultrasound (US) scanning of the spine is becoming an increasingly viable radiation-free alternative to CT scans and fluoroscopy. However, due to the complex shape of the vertebra, three-dimensional (3D) US reconstructions generated from two-dimensional (2D) US scans often lack important anatomic information, such as the spinous process. This paper investigates scanning strategies that reorient the probe during US scanning to improve surface coverage of 3D US reconstructions. A two-scan procedure with a path re-planning algorithm is presented. The proposed algorithm uses information from a first exploratory scan to generate an improved imaging trajectory whereby the US probe is near-perpendicular to the targeted bone surface. The results show a 30.4%, 42.3%, and 75.0% improvement in surface coverage on a synthetic phantom, cadaver, and human volunteers, respectively, achieving up to 56% surface coverage on human volunteers. These results emphasise the value of exploiting information about the underlying anatomy to optimise the scanning trajectory. The increased surface coverage of the 3D US reconstructions will provide higher quality radiationfree visualisation, extending the role of US as a complementary imaging modality for safe and effective diagnosis and spine interventions. © 2018 IEEE.

关键词： Motion planning

来源：评论

学校读者我要写书评

暂无评论

Privacy-Preserving Orthogonal Aggregation for Guaranteeing Gender Fairness in Federated Recommendation 25

Privacy-Preserving Orthogonal Aggregation for Guaranteeing G...

引用

18th ACM International Conference on Web Search and Data Mining, WSDM 2025

作者： Zhang, Siqing Ding, Yuchen Tang, Wei Sun, Wei Liao, Yong Zhou, Peng Yuan University of Science and Technology of China CCCD Key Lab of Ministry of Culture and Tourism Hefei China Aarhus University Department of Electrical and Computer Engineering Aarhus Denmark

ISBN: (纸本)9798400713293

Under stringent privacy constraints, whether federated recommendation systems can achieve group fairness remains an inadequately explored question. Taking gender fairness as a representative issue, we identify three phenomena in federated recommendation systems: performance difference, data imbalance, and preference disparity. We discover that the state-of-the-art methods only focus on the first phenomenon. Consequently, their imposition of inappropriate fairness constraints detrimentally affects the model training. Moreover, due to insufficient sensitive attribute protection of existing works, we can infer the gender of all users with 99.90% accuracy even with the addition of maximal noise. In this work, we propose Privacy-Preserving Orthogonal Aggregation (PPOA), which employs the secure aggregation scheme and quantization technique, to prevent the suppression of minority groups by the majority and preserve the distinct preferences for better group fairness. PPOA can assist different groups in obtaining their respective model aggregation results through a designed orthogonal mapping while keeping their attributes private. Experimental results on three real-world datasets demonstrate that PPOA enhances recommendation effectiveness for both females and males by up to 8.25% and 6.36%, respectively, with a maximum overall improvement of 7.30%, and achieves optimal fairness in most cases. Extensive ablation experiments and visualizations indicate that PPOA successfully maintains preferences for different gender groups. © 2025 Copyright held by the owner/author(s).

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：